PyTorch Foundation

A vendor-neutral foundation within the Linux Foundation that governs PyTorchPyTorch, prop. n.: The most widely used free software toolkit for building and training neural networks, especially in research; programmers write ordinary Python and it handles the calculus and the graphics chips. and, since 2025, a growing set of related open-source projects. It holds the projects' trademarks and other assets, raises money from member companies, runs the annual PyTorch Conference and community programs, and pays for shared infrastructure such as continuous-integration computing. Business governance and technical governance are separate: a governing board drawn from paying premier members oversees budget and strategy, while maintainers and contributors make technical decisions. Its founding premier members in 2022 were AMD, Amazon Web Services, Google Cloud, Meta, Microsoft Azure and NVIDIA. In 2025 it became an umbrella foundation, adding the inference engine vLLMvLLM, prop. n.: A free, open-source program for running large language models on servers efficiently, so one machine can answer many more users at once. Many companies use it to host AI models. and the training-optimization library DeepSpeedDeepSpeed, prop. n.: Free software from Microsoft for training and running AI models too big for ordinary tools, spread over many chips; it is best known for a memory-saving technique called ZeRO. in May and the distributed-computing framework RayRay, prop. n.: Free software that spreads Python programs, including AI training and model serving, across many computers and graphics chips so that big jobs run in parallel. in October, with the stated aim of hosting an open software stack for training, serving and deploying models.

Once PyTorch moved to the PyTorch Foundation, Meta became one board member among several.